There were a variety of models within the B-Series, both planned and canned.
In this section I will go over some of the specifics, but mostly concentrate on
the one that actually saw full production - the rarefied B128. Of all the models,
the American B128 is the most common.
The earliest machines were internal development prototypes. These systems,
rarest of the rare, are housed in CBM 8032 cases modified for the new keyboard
and motherboard design. All of these were hand-built in mid-1982, and sported
keyboards with black color keys. Prior to that it is logical to assume there
may have been some wire-wrapped, board-level prototypes based on hardware
development processes of the era, but none are known to exist. All such
prototypes would have had mostly ceramic packaged versions of the key ICs, all
ICs in sockets, and used EPROMs rather than ROMs for the firmware. I am only
aware of one of these prototypes surviving today.
The next most common U.S. models are the earlier version B500 and the P500 models followed by the CBM 128-80, and finally the CBM 256-80 as the rarest of all in the U.S. In Europe, the rebadged B-128 - the 610 - is the most common, with somewhat more of P500 (note the use of the same naming convention in the U.S.), B500, 710, and 720 models. This is likely due to Commodore proceeding further with sales of these systems in Europe than in North America (U.S. and Canada) whereas the US fire sale dump of the B128-80s made them the most common to find. While the first fire sales in the U.S. were during 1986 for CBM 128-80s and CBM 256-80s, these represented a limited production run (about 60 units) that had been warehoused and then unloaded when discovered after the B128-80 inventory was already liquidated. Nobody is known to have purchased one of these hi-profile systems from a dealer in North America. In Europe, all models saw some dealer promotion time - though limited - because Commodore in West Germany kept selling into 1987.
If you've read the other material available at this website, you already know that Commodores' management was in quite a churn over the B-Series product line. Not only were many engineering revisions going on, but many marketing ones as well. In looking at the historical data, it almost seems like chaos ruled during the brief lifetime of this product line.
Very rarely do you see very costly complete reversals in direction, changes in features or content after production has started and systems have begun shipping - even from Commodore. Here's an example: this is the first series users guide. Compare this to the final series users guide. Notice how in the first series edition that there was a "C128-40"? This machine was originally known as the P128, often referred to as a PET II as well, at the first public announcement but only saw limited production badged as the P500. But before the cancellation of this model, Commodore released to the North American press that the name would be the C128-40. It was the only member of the B-Series line that was to be equipped with color graphics and a 40-column display like the Commodore 64. By the time the first (and last) production series guides had been printed, all references to this system had been removed. But even when the first series edition guides were being distributed, this machine was already history. Only several dozen are known to exist of USA manufacture, all pre-production demonstrator models. It is also interesting to note that the B-Series user guides (both early and late editions) were the only Commodore users guides not formally registered for copyright with the U.S. Library of Congress or assigned a CBM part number. Clearly things in Commodore for this series fell apart very fast - even as the machines were heading to market.
Even product naming was in a churn during this time. While one can almost
guess that the rebadging for Europe had at least a little something to do with
the fact that there was an automobile product line with the same name already
(i.e. Opel Commodore B Series), it doesn't explain why the first machines out
the door were badged as B500 units instead of B128-80s or 610s. Perhaps Business
and Personal 500 series was the first thought. And what was
the significance of the "500" in the name anyway? Early information
Commodore disclosed to the trade press indicated that the number scheme was
such that x00 models had 64KB, x10 models had 128KB, and x20 models had 256KB
of RAM with "x" being a 5 or 7 initially (later a 6 as well). While
this holds true for production European models, it does not explain the B500,
P500, and the European B700 pre-production run models. The fact that many of
the B500 models came from the factory with 256KB of RAM installed doesn't lend
any clues either - but it is another interesting fact. Another logical theory
on the naming is that the European naming was originally going to be worldwide.
The naming started as the 500/700 series with the presence of lo-profile
machines only in the 500 line and BOTH lo-
and hi-profile machines in the 700 line. Later on, the 600 series seems to have
been added as a way to distinguish the business lines between lo- and
hi-profile with the 500s then being consumer only. By
then the models were further classified by the factory-installed RAM, with x10
models having 128KB and x20 models having 256KB and x00 just being a family
name with no models coming with 64KB. Looking at a mid-series European users
guide the title is 500/600/700 series. But from an American business marketing
point of view, all of these numerical naming conventions were a very lame idea.
And maybe that's how it turned out that Commodore headquarters had one naming convention
and Europe another for what was essentially the same computer series sold in
North America.
Physical development started in mid-1982 with the prototypes. At least one of
these has survived, but there certainly should be a couple of others. These
earliest incarnations were constructed within cannibalized Commodore 8032
cases. They were "hi-profile" designs and the hi-profile
"was" the B-Series. The lo-profile was a variant, not the primary
point of the design. From the example I've seen, these include ceramic 6509
CPUs and ceramic SID & DRAM chips with every IC in a socket and no ROMs,
only EPROMs. Also all pre-production prototypes would sport black keys for the
keyboard. Some of these early development prototypes would be non-functional and
used for marketing promotional materials only.
The B systems were an international effort as well. Commodore had two principle
development laboratories of point of control - the U.S. lab in West Chester, PA
and the Corby lab in the UK. All locations, whether they be marketing, sales,
manufacturing, or supporting engineering locations took their direction from
the U.S. headquarters. From internal Commodore documents and memos, I've
reconstructed the major roles and responsibilities for the B Series project (aka
CBM-II):
Overall system design and development including power/mechanical: USA
Microprocessor design (6509) and supporting ICs: MOS USA
Overall marketing requirements: USA
Core engineering development including motherboard, BASIC and KERNAL firmware:
USA
Internal disk drive option (8250LP derivative), printer options: Japan
Board layout and design for all motherboards worldwide: USA
Integral display design and manufacturing worldwide: Japan
Programming documentation for software developers: USA (until December 1983),UK
(1984 till abort)
Product demonstration software: UK
Licensed software titles: USA and UK (depending on their marketing
requirements)
8088 coprocessor board option: USA
MS-DOS and CP/M-86 porting: Microsoft USA (MS-DOS), CBM USA (CP/M-86)
End User documentation: USA (UK handled localization to both UK and German)
Motherboard manufacturing, final assembly, distribution, and service: North
America (USA), Europe (West Germany)
European sales outside UK: West Germany
The product plan at the final stage was to be as follows, where all models were
80-column displays and the "X" in the model indicated a built-in
coprocessor:
|
B128-80 (610 in Europe) |
BX128-80 |
|
B256-80 (620 in Europe) |
BX256-80 (630 in Europe) |
|
CBM 128-80 (710 in Europe) |
CBMX 128-80 |
|
CBM 256-80 (720 in Europe) |
CBMX 256-80 (730 in Europe) |
The "B" alone indicated a lo-profile model, whereas the
"CBM" in the name indicated a hi-profile model. All CBM models were
to offer optional internal dual diskette drives that were most similar to 8250
LP assemblies, but only prototypes were ever produced and these in Europe only.
No known X models ever reached ship-level production. All but the B256-80 saw
some production of the non-X models. Follows is a review of all the B-Series
systems made and when for completeness.
There was only one of the consumer models of the 500 series ever produced. The
Commodore P500 saw limited production at both U.S. and West German plants in
very late 1982 thru early 1983. These units were most probably for the purpose
of ironing out the manufacturing process and had officially been announced as
dropped by the Summer 1983 CES show. The P500 is a
lo-profile B-Series model with 64KB of BASIC RAM (plus an additional 64KB of
RAM shared with the VIC II chip only for a total of 128KB) and sharing the
basic features of the other members of the series. The most notable unique
features were two joystick ports, running the 6509 at 1MHz instead of 2MHz, and
sharing the same video chip as the Commodore 64 in support of its intended use
as a consumer model. It was never put into general production and is documented
as the first model dropped in the series in favor of the Commodore 64 model.
The engineering change order (ECO) for the P500 cancellation was recorded on
March 11th, 1983.
It is ironic, though, that the P500 - the one that was first to be dropped -
would end up being the only one to survive in some form. After the demise of
the B Series, another system along the same hardware architecture was
envisioned, known as the D128. The concept became, much later in early 1986,
that Commodore would again launch a 128K computer; this time with compatibility
with the Commodore 64, working CP/M®
support via a second processor (Zilog Z80) as well as
a diskette drive peripheral capable of reading non-Commodore MFM encoded
diskette formats - the 1571. The production Commodore 128 (C128) ultimately
contained much of the technology and experience learned in producing the B-Series
and P500, and attempting to market them along with the conceptual D128
prototype ideas.
There were two of the business lo-profile models of the 500 series badging. The
Commodore B500 saw limited product in both U.S. and West German plants in late
1982 thru early 1983. These units too were most probably for the purpose of
ironing out the manufacturing process and in support of early software
development. The B500 is a lo-profile B-Series model with either 128KB or 256KB
of RAM (there are factory made examples of both known to exist) and sharing the
same basic features of the other members of the series but installed with only
the 128K version BASIC ROMs. The most notable differences from later production
models were the layout of the motherboard similar to that of to the P500 model
(e.g. bottom mounted power supply on motherboard risers).
Later production and using the final naming conventions the Commodore B128-80
and 610 saw production in both the U.S. and West German plants in 1983 thru
project cancellation in mid-1984. These units were all production run models
sold to end users. They are both the same computer, only different in name
badging, power supply, video standard support (e.g., NTSC or PAL), and plant of
manufacture. No model B256-80 systems are known to exist, but I have personally
examined a lone factory sticker labeled Commodore 620 system in a private
collection here in the U.S. that was brought over from Corby by an officer of
CBUG and saved from destruction. It is a factory-made model, but probably not
production and more like the B500/P500 runs. The B128/610 lo-profile model is
the most ubiquitous of the B series. It sets the standard for features in the
series.
In terms of 700 series models, the Commodore CBM 128-80 saw production in both
the U.S. in mid-1983 only, the 710 in West Germany from mid-1983 thru late
1984. Some prototype units were made in late 1982 (and placed in CBM 8032 cases) thru very early 1983
(badged as B700 or B715 units in Europe only). These are notable because the top half of
the case has the blue painted Commodore logo molded in like a lo-profile unit
which disappeared in later production. They may also have a modified
lo-profile motherboard (cut and adapted to fit the hi-profile case)! But the CBM 128-80 and 710 badged units
were the only production run models, though limited in volume - particularly in
North America. They are both the same computer, only different in name badging,
power supply and related video jumpers, and plant of manufacture. All of these
models are hi-profile units with 128KB of RAM and sharing the same basic
features of the other members of the series.
The Commodore CBM 256-80 saw production in the U.S. in mid-1983 only, the 720
in West Germany from mid-1983 thru late 1984. These units were production run
models, though limited in volume - particularly in North America. Again, they
are both the same computer - just different as the CBM 128-80 is to the 710.
This is also a hi-profile model but with 256KB of RAM and thusly having the
notable difference of unique BASIC ROM code.
The B-Series actually consisted, in the end, of the B128-80 as the only production level low-profile system (i.e. without integrated display), and the CBM 128-80, and CBM 256-80 hi-profile systems here in the United States and Canada as the only production hi-profile systems. In Europe, the 610 - as the rebadged B128-80 - as the only production low-profile system, and the 710 and 720 as the rebadged CBM 128-80 and CBM 256-80 respectively. In the UK, several examples of the CBMX 256-80 (aka 730) were discovered by CBUG members during the Corby plant closure visit, but these were not production models. They sported internal dual-diskette 8250LP-type drives (ala the later CBM 8296D models) and 256K RAM with non-standard power supplies (so the drives would fit!). Several of these development systems also sported the ultra rare 8088 co-processor boards. One example of the Commodore 620 is also known to still exist as mentioned previously (a model that would have been known as the B256-80 here in North America) but these too are not likely to be production models. Since 1998 I have captured the serial numbers of B128-80 production models sold on eBay to give a representation of the number of units manufactured in the USA.
|
Lowest B128-80 Serial Number Recorded |
Highest B128-80 Serial Number Recorded |
|
C001495 |
C010929 |
All production B-Series models exhibited the same basic architectures around the MOS 6509 CPU, with the only significant differences being that the hi-profile models had a different character generator ROM which provided for a better character set to take advantage of the integral display and a different set of BASIC ROMs in the case of the 256KB RAM model. Another interesting difference is that the P500 model sports not one, but two N82S100 FPLAs! But other differences did exist even in production, the only ROMs which ended up being able to support the commercial software titles that did get released were the 128KB RAM model version. The 256KB RAM models would not run the few commercial titles released in the U.S. due to BASIC ROM differences. Even more evidence of the chaos preceding the aborted launch.
The B-Series, as a family, were the most advanced 8-bit Commodore computers ever constructed at the time. Conceptually, probably even the most advanced 8-bit computer ever constructed by any computer company given the Intel 8088 coprocessor, one megabyte RAM expansion, MS-DOS compatibility, CP/M-86 compatibility, integral display, IEEE-488 parallel port, and integral 8250LP one megabyte storage capacity. These original features dwarf the later, and more well-known, C128s capabilities from an architectural perspective. Nothing then or since has come close to the advanced capabilities, durability, and expansion potential of this design around an 8-bit microprocessor with optional additional 8-bit and 16-bit processor expansion options. Just witness the fact that Commodore based all their follow-on 8-bit products along design lines and architectures first established with the B-Series.
In the production versions it offered the following features to U.S. customers along with a 90-day factory warranty:
When you opened the box, it included a users guide and grey AC power cord, both hi-profile and lo-profile systems, and a 5-pin DIN to one RCA jack video cable on lo-profile systems only. Obviously hi-profile systems also sported an integral 80 x 25 green monochrome display and a detachable keyboard. The P500 models required an accessory 8-pin DIN to three RCA jack video cable just like the later design (not 5pin DIN video) Commodore 64 would use.
This table summarizes the basic hardware architecture of the B-Series systems:
|
Microprocessor |
MOS 6509A* |
$0000-$0001 |
|
Video |
MOS 6545A-1** |
$D800-$D900 |
|
Audio |
MOS 6581 |
$DA00-$DB00 |
|
Keyboard |
MOS 6525A |
$DF00-$E000 |
This table summarizes the basic I/O architecture of the B-Series systems:
|
RS-232C Port |
MOS 6551A |
$DD00-$DE00 |
|
IEEE-488 Port |
MOS 6525A |
$DE00-$DF00 |
|
Datassette™ Port |
MOS 6525A |
N/A*** |
|
User Port |
MOS 6526A |
$DC00-$DD00 |
* The P500 used the 6509 and not the 6509A due to
it being clocked at only 1Mhtz though the faster rated part will service the
P500.
** Most models use the Motorola 68B45 instead of the MOS 6545A-1.
The MOS part in the required 2MHz rating was quite rare by 1984 but they are
pin compatible.
***The KERNAL ROM support for the Commodore DatassetteTM was dropped in favor of a
built-in machine language monitor on machines built after April 1983. Only the
-01 revision ROMs contain the code support to use this peripheral. All newer
revisions do not. To access the built-in monitor type "BANK15:SYS 60950" at the BASIC ready prompt. Type SYS 6 on
P500 models. Commands and usage are covered in the final B series users guide.
The backside of the machines consisted of the following ports: IEEE-488 PET card edge, Datassette, RS-232C, 1/8" mini jack audio out, 5-pin DIN video out (driving either composite or TTL monitors), expansion cartridge and an IEC power connector. There was also a reset button, an IEEE-488 ground connection, and, on very late production B128-80 models only - a fuse holder accessible from the rear as well.
The P500 differs from the other lo-profiles in that it has an 8-pin DIN video out (same as the later C64 design), two joystick ports in addition to the other ports and an RCA audio output connector instead of the mini jack on the other models (though audio is also provided thru the 8-pin DIN connector also ala C64). Also the P500 has 40-column color video capability so does not use the Motorola 68B45 but instead the MOS 6567 (MOS 6569 for PAL) just like the original Commodore 64.
There are several differences in production firmware and hardware. Here are some of the things to look for:
First, let's establish how to identify a lo-profile B computer. Here is a view from the top, the bottom, and the rear left. rear middle, and rear right for detail. This is pretty much how they all looked like, from the B500 to the B128-80 and even the European 610. Externally the only obvious difference is the B128-80 has two of the top vents closed as there is a plastic mount for the power supply here. Other models have all the vents open. The P500 had the most notable differences. There are two joystick ports unique to the P500. And of course the P500 had differences for the video and associated video RAM. The P500 uses an RCA connector for auxiliary audio, not a mini-jack. And the P500 has no built-in speaker. Another obvious difference on the P500 is that it is the only model to have a green power LED. All the other models have a red power LED, though this is probably more due to the early production stage of the P500 than anything else.
But there is one external difference to look for on very late production units - at least in the United States produced models. Somewhere between B128-80 serials C010300 and C010800 a fuse, and associated label, was added to the power supply harness accessible from the backplane. This change was very similar to other commodore products of the time with integral power supplies such as the SFD-1001, 2031LP, and 1541 diskette drives. This change is a good way to immediately tell that the unit you are looking at is one of the very last USA made B-128s ever produced.
When you power-on the system unit, you will get the following banner message on 128K RAM machines:
*** commodore basic 128, v4.0 ***
On factory-equipped 256K RAM models you will get this power-on banner:
*** commodore basic 256, v4.0 ***
This power-on message is the same regardless of whether this is a low-profile or hi-profile model, and is based totally on the original factory installed RAM configuration since the ROMs were different depending on this only. If you add an additional 128KB RAM to a 128KB RAM system to get to 256KB RAM total, the power-on banner will stay the 128KB version, for example. The hi-profile models have a superior character set ROM, an 8 x 16 matrix that takes advantage of the integral green phosphor CRT monitor. The low-profile model has - regardless of the factory RAM configuration - a character set ROM that is compromised for composite video output, an 8 x 8 matrix, and is less sharp on higher end monitors comparable to the integral monitor on the hi-profile models. The output of the video will be the color of the monochrome monitor CRT with the low-profile "B" models, either amber or green or white on black - if connected to a composite video color monitor like the Commodore 1702. In fact, the pin outs for the composite video feed and ground are compatible to the Commodore VIC-20 or very early Commodore 64 models (those with a 5-pin DIN video connector) and you can use - or make - the same type of 5-pin DIN to 2-RCA jack cable as for the VIC-20 if you do not have the original cable that came with the low-profile model B computer. However, do not plug in the audio jack lead of the cable if doing this as there is no audio output within the B-Series video connector, there is a separate mini-jack connector on the B-Series for audio output. The hi-profile models have no optional video output, having only a pin connector internally on the motherboard for the integral CRT and are thusly only green in color. The P500 models, being 40-column color, came up in a color power-on screen color scheme very similar to the VIC-20, and would use the same cable that was originally included with the Commodore 1701 and 1702 monitors. But the message was the same as the other 128KB RAM models. Again the P500 model had an 8-pin DIN video connector that was exactly the same as later model Commodore 64s, but could also take a 5-pin DIN video connector just like the Commodore 64 models.
To identify things further you will have to open up the system unit. Inside you will find most of the key chips and ROMs are professionally socketed by the factory for ease of servicing. Both early and final production firmware for the BASIC interpreter was for version 4.0 of Commodore’s implementation of Microsoft® BASIC. Internal Commodore engineering memos indicate a revision 04 of parts 901240, 901241, 901242, 901243 and 901244 were developed to fix problems in the PRINT USING command and some more minor BASIC bugs on September 29th, 1983; but they were never officially released via ECO due to the project cancellation by 1984. The hi-profile level of the FPLA is the only one to support the 8088 co-processor option reliably, 906114-05 (the lo-profile model suffered a design defect of an unterminated DRAM data bus as well as insufficient power supply that precluded use of the 8088 board reliably as well). Note that the P500’s two-FPLA design is not compatible with those used in either the other B Series systems or the Commodore 64 (p/n 906114-01). And its logic equations were never corrected to support the 8088 board. While they all have a similar part number, they are all very incompatible, with the "revision" number really only being a model identification number with -02/03, -04, -05 for the various B Series models. This is not unlike the PLA numbering scheme used later in the 8296/8296-D, Commodore 16, Commodore Plus/4 and even the 1551 where the PLA was a MOS clone of the Signetics N82S100 known as the 8700 and used 8700-009 thru 011 as model identification numbers. The key B-Series firmware ROMs and part numbers that reflect the information described are shown below along with many downloadable JEDEC format images for blowing fuses with an FPLA programmer and absolute binary format images suitable for N82S100 programming:
|
Commodore B Series ROM & FPLA Sources |
|
BASIC Hi (Non-P 128K Models) |
901242-03 |
|||
|
BASIC Lo (Non-P 128K Models) |
901243-03 |
|
BASIC Hi (Non-P 256K Models) |
901240-01 |
901240-02 |
||
|
BASIC Lo (Non-P 256K Models) |
901241-01 |
901241-02 |
901241-04 |
|
KERNAL (P500 Model Only) |
||
|
BASIC Hi (P500 Model Only) |
||
|
BASIC Lo (P500 Model Only) |
|
English Character ROM (Lo-Profile Models) |
|
|
English Character ROM (Hi-Profile Models) |
|
|
English Character ROM (C64/P500 Models) |
|
Co-processor BIOS ROM (All Models) |
These tables represent part numbers from the last USA factory-installed
versions of these parts as well as the last field service levels from Commodore
known. This includes the unreleased -04 version announced in a July 16th, 1984
internal memo. If you have something newer with an official Commodore
silk-screened part number on the ROM or even an official -04 silkscreened ROM
(which should not exist) let me know about it. . When making replacements using
EPROMs, note that all the above are either 2364 mask ROM images or 2764 EPROM
images, except for the coprocessor image which is a Texas Instruments TMS2532
EPROM. You can use a Motorola MCM68764C or MCM68766C EPROM, or even a Texas
Instruments TMS2564 EPROM (just let the top pins hang over the socket), to
replace the original 2364 ROMs. If you cannot locate one of these Motorola
EPROMs you can construct an adapter for a standard 2764 EPROM just like Commodore
did in later production. On the 8088 board you can substitute a TMS2532A EPROM
if necessary.
You can blow a replacement FPLA using either the Signetics
82S100N or the Phillips PLS100N. You may also encounter the Mullard
N82S100. Another compatible part is the Fairchild 93459 which is a pin
compatible substitute for the N82S100. All are equivalent parts. But you must obtain a new blank part, unlike
an EPROM an FPLA cannot be erased or re-used. They are "write once"
devices. You cannot re-create these with an EPROM programmer, you will need a
more advanced logic programmer such as the Data I/O 2900, 3900, 3980, or
3980-XPi. The download links in the table above are to the JEDEC (JED file
extension) fuse blow map images for each revision of the FPLA parts in the P
and B series. Each was made and tested using a Commodore USA factory installed
original part.
*Note: The P500 was unique among Commodore designs in that it had not one, but
two FPLAs. As stated earlier these are not revision number parts, you need
both. For trivia, the 906114-01 was also not a revision number, this is the PLA
used in the model Commodore 64! Commodore part 906114 is simply their number to refer to the PLA device.
Follow these steps to create a 28pin to 24pin EPROM adapter:
1) Obtain a 28pin DIP socket and connect pins 28, 27, 26 and 1 together using
30 gauge wire.
2) Connect pin 20 to pin 14.
3) Obtain a 24pin DIP socket and connect pin 18 of this socket to pin 23 of the
28pin socket.
4) Connect pin 21 of the 24pin socket to pin 2 of the 28pin socket.
5) Bend down pin 23 of the 28pin socket and insulate so it will not make
contact when inserted into the 24pin socket.
6) Bend down pin 20 of the 28pin socket and insulate so it will not make
contact when inserted into the 24pin socket.
7) Solder all wire connections.
7) Insert the 28pin socket into the 24pin socket lining up pins 14 and 15 of
the 28pin with pins 12 and 13 of the 24pin socket (i.e., bottom ends lined up).
8) Insert your programmed 28pin EPROM into the 28pin
socket and insert the whole assembly into the appropriate 24pin socket on the
motherboard.
If you do not have an EPROM programmer to save your ROM images then you can save to your CBM diskette drive using these steps:
1) Enter TIM, the built-in machine language monitor, by typing "BANK15:SYS 60950" at the BASIC ready prompt (SYS 0 for P500 models). You should see the banner ** monitor 1.0 ** and a register dump now along with a "." prompt.
2) Now at the monitor "." prompt type the following S"0:KERNAL",08,0FE000,0FFFFF and then press return. This will save the image of your KERNAL ROM to drive 0 of your IEEE-488 diskette drive.
3) Now to save the BASIC ROM image type the following at the monitors' "." prompt S"0:BASIC",08,0F8000,0FC000 and then press return again.
4) To exit the monitor type X at the "." prompt. You will be returned back to the BASIC 4 interpreter.
As a reference, here is a short table of the commands available while in the B-Series Terminal Interface Monitor (TIM) program:
|
M |
Display memory |
|
: |
Alter memory |
|
R |
Display registers |
|
; |
Alter registers |
|
G |
Begin execution |
|
L |
Load |
|
S |
Save |
|
V |
View Segment |
|
U |
Set default disk unit |
|
@ |
Send disk command or get disk status |
|
X |
Exit to BASIC |
|
Z |
Transfer to second microprocessor (e.g. 8088) |
|
<filename> |
Load and execute |
Now here's a view inside the B128. Here you can see where the factory would
add the additional 128K
of RAM to make a 256K RAM model. Nearly all production uses Mostek
brand 4564-15 DRAM chips. This is the same motherboard used for all production
lo-profile machines (including the European 610 and 620 models). Early
non-production models like the B500 and P500 had
some differences. The most obvious difference is that the power supply on these
early models was mounted on the lower half of the case, near the motherboard
instead of the top half of the case as on later production units. Other
differences of note include either ceramic 24-pin ROMs (very early production)
or 28-pin EPROMs on 24-pin DIP (very late production) adapters.
Remember that in 1983 the 64K RAM chip was the standard hence the use of that
configuration for the memory expansion. And here's a view of those ROM chips that you just read
about earlier. And the P500 ROM setup showing the EPROM adapter daughter board.
Finally, here's a view of the heart of it all, the MOS 6509A microprocessor and
next to it an Hitachi sourced version of
the Motorola 68B45 video chip. The P500 was 1Mhtz, but it used the MOS 6569 as
the Commodore 64 did - so it was another animal altogether.
The views inside the European model 610 and 620 are identical, only the
differences that control the video between the NTSC standard used in the United
States and the PAL standard used in Europe can be found on close inspection.
The view inside the hi-profile models has obvious
differences in cabling and board layout, of course, due to the difference
in power-mechanical design. But the motherboard componentry is essentially the
same. Here, the only relevant difference is the different firmware ROMs and
lack of an external
video connector. The most obvious external differences are, of course, the
integrated green phosphor monochrome display as well as the detached
keyboard. The hi-profile models also have a cooling fan installed on their
power supplies which the lo-profile models do not (though they could sure use
it). You can raise the "hood" on the hi-profile models by removing
the two machine screws that attach the top half of the case to the back plane.
Once you raise this the speaker cable will have to be disconnected and possibly
the monitor cable in order to fully raise the top. There is a "hood
prop" on the right side that was provided to prop the top open for service
operations.
Do not bother to look for the Z80 coprocessor board; it was officially
dropped in early 1983 from internal Commodore memos I have. But if you happen
to be lucky enough to find one of the ultra-rare Intel 8088 co-processor
boards, when you look inside then you would be interested to know that the last
engineering revision information was part number 326234 and artwork 326236 rev
F. These are the final board revisions that Commodore was ever known to have
before the project was cancelled. The engineering design was rather
interesting. Both processors were always active. The 6509 CPU handled I/O
(e.g., keyboard, IEEE-488, user port, video, etc) and
the 8088 handles all the computations and data manipulation. Noting the JPEGs
of the coprocessor board in links on this page note the MOS interface chips in
their usual role of interfacing, but doing so with Intel processor technology.
The Commodore designed prototype co-processor boards will only work in the hi-profile
models due to PLA, power, cooling, and physical space restraints indicative
of the aborted engineering efforts. There was a third party company working
thru the CBUG making an NEC V20-based board that would fit in and be power draw
compatible with the lo-profile units as well, but there is no evidence this got
past the prototype stages either. In practice the Commodore engineering boards
only had AMD 8088 processors installed. If you
have a version of a Commodore-made coprocessor with an Intel 8088 or other than
artwork rev F on the silkscreen or a known third party design by Anderson
Communications Engineering which was
advertised to use the NEC V20 co-processor
board for the B-Series please e-mail me!
Either of these two co-processor designs could also carry an Intel 8087 math
co-processor as well, but flaws in the board layout prevent their use and
programs that attempted to access them would hang the computer. If you install
the 8087 and the try to boot the MS-DOS 1.25 diskette, it will also hang the
computer for the same reason. I have debugged the problem if you are interested
write me for the cuts/adds to make to the board for it to now work.
A set of preliminary schematics are contained within the B128 Programmers Guide published by Protecto Enterprises mentioned earlier. The complete final schematics are contained with the Commodore Service Manual for the B Computer; part number 314010-06 published January of 1985. This was the last known revision of the North American service manual.
Of interest are the generally available parts listings. These can help you repair your B-Series computer. Unfortunately, Commodore only documented the B128-80 as it was the only high-volume model to be released. But many of the parts were standard across the B-Series line. There are two pages which cover ICs, transitors and diodes and another which covers capacitors, resistors, general connectors and miscellaneous parts. As a service to the Commodore community I offer for sale replacement FPLA chips (i.e., 906114-XX) at parts cost plus shipping only already programmed for the version you need. I also offer EPROM duplication using the exact same pin for pin compatible parts used originally or their equivalent also for parts cost plus shipping only.
© Copyright Edward D. Shockley, 1998-2021
All Rights Reserved
